In data interfaces between two data processors, it is known to employ a buffer storage mechanism for purposes of buffering the flow of data from one processor to the other and vice versa. A block of data to be transferred is moved from one processor and stored into the buffer storage mechanism. Thereafter, this block of data is read out of the buffer storage mechanism and transferred to the second data processor. These transfers into and out of the buffer storage mechanism cannot occur simultaneously because all of the arriving data must be present in the buffer storage before the buffer emptying process can begin. Processor-to-processor data transfer methods which employ this buffer storage technique are sometimes referred to as "store and forward" methods or "mailbox" methods.
While the known methods are generally satisfactory, there is nevertheless room for improvement. In particular, it is desirable to provide processor-to-processor interfaces which enable faster data transfer rates between t e two processors. It is further desirable to provide now and improved interface mechanisms which enables multiple independent processor-to-processor operations to occur in a concurrent manner.